366 REPORTS ON THE STATE OF SCIENCE.—1912. 
Recently, on the grounds of the behaviour of worked and unworked 
metals when subjected to tensile stresses at high temperatures, the 
hypothesis has been proposed '** that a small quantity of the amor- 
phous modification persists to much higher temperatures, the ‘ tempera- 
ture of complete recuperation’ being fixed at 650° for copper and 
710° for an alloy of copper and nickel. Even this modified hypothesis, 
however, only assumes the presence of a small quantity of intercrys- 
talline material in the amorphous state, and the phenomena observed 
are in no way to be explained by diffusion through these small masses. 
An argument against diffusion in crystals is not, therefore, to be based 
on the existence of amorphous modifications of metals. 
The Electrolysis of Crystalline Solids. 
The electrical conductivity of metallic sulphides has repeatedly been 
the object of study, and it was shown by W. Hittorf?** that the 
conductivity of silver and cuprous sulphides was electrolytic in char- 
acter. The conductivity of most sulphides is, however, metallic. The 
electrolysis of natural quartz crystals was observed by E. Warburg 
and F. Tegetmeier,’** and it was found that electrolysis was only 
obtained in the direction of the principal axis. The actual electrolyte 
is the alkali silicate always present in natural quartz. Further inves- 
tigations show that the conductivity of sulphides is more electrolytic 
at high temperatures and metallic at low temperatures,'** whilst pure 
oxides of most metals conduct almost exclusively metallically.**’ 
Elaborate investigations of a large number of compounds show that 
whilst electrolysis undoubtedly occurs, it is the exception rather than 
the rule, and that solid electrolytes are usually of a saline character.*?* 
Special attention has been given to the conductivity of the mixtures 
of rare earth oxides used in the Nernst filament. These are regarded 
as electrolytes conveying an unusually large residual current.’*? It 
is possible that these masses are not entirely crystalline. 
Solid barium chloride is an electrolyte, obeying Faraday’s law,’*° 
whilst sodium hydroxide presents anomalies, indicating that at 230° 
traces of a liquid substance are present in the pores. Porcelain, which 
is only in part crystalline, is an electrolyte from 300° upwards, its 
behaviour being that of a solution of alkali silicates in aluminium 
silicate as a non-conducting solvent.**? 
Special interest attaches to the crystallised silver halides. Their 
conductivity in the solid state was investigated by F. Kohlrausch.**? 
Tt was then found '*3 that solid silver iodide is an electrolyte, and that 
13 G. D. Bengough, Jour. Inst. Metals, 1912, 7, 123. 
124 Pogg. Ann., 1851 [iii.], 24, 1. 
12 Ann. Physik., 1888 [ii.], 85, 455. 
126 R, von Hasslinger, Monatsh., 1907, 28, 239. 
127 F, Horton, Phil. Mag., 1906 [vi.], 11, 505. 
128 J, Koenigsberger, Zeitsch. Hlektrochem., 1909, 15, 97; C. Doelter, ibid., 1908 
14, 552; Monatsh., 1910, 81, 493. 
129 W. Nernst, Zeitsch. Elektrochem., 1899, 6, 41; 1900, 7, 373. 
130 #. Haber and 8. Tolloczko, Zeitsch. Anorg. Chem., 1904, 41, 407. 
13l F, Haber, ibid., 1908, 57, 154. 
182 Ann. Physik., 1882 [ii.], 17, -642. 
138 O, Lehmann, bid., 1889 [ii.], 88, 396. 
